Race (black-white) and sex inequalities in tooth loss: A population-based study.

The effect of health inequalities is determined by different socioeconomic, sex, and race conditions. This study aimed to analyze the association of tooth loss with race (defined by self-reported skin color) and sex. Based on the hypothesis that the association between tooth loss and race may be modified by sex, we also aimed to evaluate possible interactions between race and sex in association with this event, in a population-based study in the city of Campinas, Brazil. A directed acyclic graph was used to select covariates. The prevalence, of tooth loss was 19% higher in black women compared to white men (Prevalence ratio [PR]: 1.19; 95%CI: 1.05-1.34). Moreover, the prevalence of tooth loss in black women was 26% higher than in white women (PR: 1.26; 95%CI: 1.13-1.42); and, within the strata of black people, black women had 14% higher dental loss (PR: 1.14; 95%CI: 1.02-1.27) compared to black men. This study found a significant interaction between race and sex in tooth loss, with a disadvantage for black women. In addition, this work contributes to the discussion of health inequities and can support policies for the provision of universal dental care.

Introduction

The impact that social inequalities place on health is often evidenced in research and academic discussions [1, 2]. The effect of these inequalities crosses several dimensions of health and is determined by different conditions, including socioeconomic [3, 4], sex [5], and race disparities [6]. Socioeconomic status is a strong consequence of racial inequalities [7-9], since blacks individuals have a lower educational level, lower income, and tend to live in places of high social vulnerability [10, 11].

In addition to economic issues, it is necessary to consider other disadvantages, which tend to remain in various dimensions of life, even after black slavery was abolished in Brazil, which lasted about 300 years. Racial minorities, in this case black people, may biologically embody the effects of racism [7, 12], with every day (or even less common) discriminatory exposures [12]. Adversities throughout life, such as poverty, psychosocial stresses, stereotypes, and the context of housing, can affect the physical and mental health, altering cardiocirculatory, metabolic and immunological functions [8]. In oral health is possible that inequities are due to poverty [10], levels of education [4] or discrimination in health care [13]. In Brazil, skin color has been studied as proxy of race and this indicator is commonly referred as color/race [14-16]. In health, in contrast to the only biologic approach, the race may be considered as a concept socially constructed by historic dynamics and power relations [8, 12].

The literature has pointed out worse oral health conditions among black individuals compared to whites. The highest prevalence, among black individuals, of untreated caries [17, 18], periodontal diseases [19], greater need for prosthesis, difficulty in accessing dentists [6], and the influence of race on the decision of dental mutilation [13] are examples of racial inequalities in oral health in Brazil. Race is associated with the difficulty in accessing health care [20], especially oral health care. Since it is a service with higher costs due to the high prevalence and recurrent cumulative nature of caries and periodontal disease, the oral cavity is even considered the most expensive part of the body to treat [21, 22]. In this sense, socioeconomic level and monthly income, for which black individuals are highly vulnerable [17], are determining factors in the use of oral health services [4, 23].

Tooth loss is among the conditions that have most impacted the health of the world population in the past two decades. In other countries, ethnic differences were found in a sample in East London, with disadvantages for black individuals for loss of posterior tooth [24], and in Brazil, evidence of more frequent tooth loss was found in black individuals when compared to white individuals [6, 25].

Attention to sex inequalities is also required, considering the historical aspects of oppression of women, which persist today, especially regarding the situation of work, income, double shifts, and violence [26, 27]. These issues seem to have health effects, especially on emotional aspects [27], but a higher occurrence of diseases [27] and worse self-rated health [3] has also been shown for women. Nevertheless, in other dimensions, such as worse health risk behaviors and early mortality, the scenario is unfavorable for men [5]. Thus, women seem to be more vulnerable to health conditions that are more limiting and chronic, while men present more behavioral and lethal problems. Research has shown an important impact on oral conditions due to the fact of being female [28-30]. Other studies have found a disadvantage for women concerning chewing, communication, and pain difficulties [31], negative self-assessment of oral health [32], and tooth loss [3]. However, there is no consensus in the literature, and there is one study reporting the absence of gender differences [33].

The evidence of racial and sex inequalities in tooth loss could bring the subgroup of black women underprivileged in relation to this theme. In this way, it is worth considering the analytical strategy of intersectionality, developed by Crenshaw in 1989 to study the black feminism. Intersectionality, drawing on a set of social issues and perspectives, explains an experience of multiple subordination that cannot be reduced to the simple sum of underprivileged conditions [34].

This study aimed to investigate a possible effect and interactions between race and sex on tooth loss, in a population-based study. To our knowledge, there are no studies looking at the relationship between race and sex in relation to tooth loss and this study contributes to filling this gap.

Methods

Study design and sampling

This research is a cross-sectional, population-based study, conducted with data from the Health Survey of the Municipality of Campinas (ISACamp) in Brazil.

A stratified probabilistic sampling was carried out by clusters and in two stages: census tract and household. In the first stage, 70 census tracts were drawn with probability proportional to the size (number of households). 14 census sectors were drawn from each of the five existing Health Districts in the municipality. In the second stage, households were selected by a systematic draw applied to the list of households existing in each of the sectors drawn [35].

A minimum sample size of 1000 individuals was defined for the age domains of adolescents (10 to 19 years old) and elderly (60 years or over) and 1400 for adults (20 to 59 years old). To reach this sample size, after updating the address list of the selected sectors in the field, the numbers of households for each age domain were independently selected. The definition of the number of households that should be drawn was based on the expected average number of people per household (people/household ratio) for each district, based on the 2010 Census data. The desired sample size (1000 or 1400 people) was divided by the corresponding people/household ratio. However, considering the presence of non-responses, larger numbers of households were drawn, considering non-response rates from previous surveys. Data were collected from 3021 people aged 10 or over [36], using a pre-coded questionnaire, applied through tablets, by interviewers trained.

Study variables

The outcome variable was self-referred tooth loss, extracted of the question: “Have you ever lost a tooth (upper or lower)? If so, did he lose one, more than one or all of his teeth? (Disregarding extracted teeth to place braces, wisdom and milk teeth)”, with the options to response: (1) No, (2) yes, only one tooth, (3) yes, more than one tooth, (4) yes, all teeth. The variable was analyzed in two ways: 1) as a dichotomous variable (loss of at least one tooth or none); and 2) as an ordinal variable classifying each participant into any of the following four categories: No loss, loss of 1 tooth, loss of more than one tooth (but not all), or loss of all teeth. When asking about tooth loss, the interviewers informed the respondents to exclude teeth extracted for orthodontic reasons, disregarding teeth extracted for braces, wisdom and baby teeth. Tooth loss is an important marker of oral health, since each tooth is a dental organ and its subtraction has consequences on occlusal adjustment, temporomandibular joint disorder, reduction of the masticatory board and consequently overload for the digestive system [37]. Besides, the loss of at least one tooth since it represents neglect in the dental field, resulting in an increase in the level of severity of oral diseases, in addition to reflecting the model of oral health care adopted and the way individuals understand the condition [38]. Therefore, we decided to focus the primary analysis on the dichotomous outcome, which also represents the most practical way to assess and present interactions [39, 40]. However, on the other hand, the analysis of the same variable as ordinal can show us a gradient of tooth loss [41].

The race variable studied was self-reported. In Brazil this is the official method of racial classification since 1991, based on the individuals’ own perception of the color of their skin. In the present study, race/skin color variable was denominated “race” and categorized as black, brown, and white individuals. It was verified that the categories of black and brown presented a similar association for tooth loss, in each sex and in each age, so these two categories were joined to compose the group of black individuals (also used by IBGE-Brazilian Institute of Geography and Statistics) [42]. Individuals who declared themselves as yellow, indigenous, or other races were not studied, as they represented less than 2% of the population.

In this research, the application of directed acyclic graphs (DAG) was used [43] to identify the factors that would need to be conditioned to control confounding. DAGs have been increasingly used as a tool to guide the design and analysis of epidemiological studies [43]. These graphshelp guide the adjustment to a specific set of variables, avoiding problems such as collision bias, overfitting, and unnecessary adjustments [44, 45].

The DAG included the following measured socioeconomic and demographic variables: age, schooling in years of study (0–4, 5–8, and 9 or more years), monthly income per capita in minimum wages (MW), district of residence (East, North, Northwest, Southwest, and South), last visit to the dentist (less than 6 months, between 6 months and less than 1 year, between 1 year and less than 2 years, and 2 years or more or never consulted). Other variables such as, health behavior, patterns of participation (individuals who answered the survey because they spend more time at home, such as women, the elderly, and children), and familiarity (characteristics of the family, such as housing and cultural regions), were not measured in the study were included in the DAG. to assess the need to conditioning on them in the multiple models carried out.

The Vulnerability Index of the State of São Paulo (IPVS) was also included, referring to each sector included in the survey, stratified as the most very low, very low, low, average, high, and very high vulnerability.

The IPVS consists of a typology of situations of exposure to vulnerability, adding to the income indicators others referring to the family life cycle and schooling, in the intra-urban space. It is composed of socioeconomic variables (income and literacy status), as well as demographic variables related to the family life cycle (presence of younger children, age and sex of the head of the family). The IPVS is conceptually composed of two assumptions, the first being the finding that the numerous dimensions of poverty need to be considered in a study on social vulnerability. In this sense, the IPVS operationalizes the concept of social vulnerability [46] that the vulnerability of an individual, family or social group refers to their greater or lesser capacity to control the forces that affect their well-being. Thus, vulnerability to poverty is not limited to considering income deprivation, but also family composition, health conditions and access to medical services, access and quality of the educational system, the possibility of obtaining quality work and adequate remuneration, the existence of legal and political guarantees. The second assumption on which the IPVS is based is the consideration that spatial segregation is a phenomenon present in State of São Paulo’s urban centers and that it contributes decisively to the permanence of patterns of social inequality.

In the DAG, we represented the effects of race and sex on oral health, which can be mediated by income [3, 4], education [4] and, more proximally to the outcome, by health behaviors and the dental care [6]. Other variables, such as age, familiarity, patterns of participation in the research, place of residence and IPVS [7], were also considered in the causal diagram (Fig 1). Because it can be a list of unmeasured common causes of both place and race, we follow a convention that visually simplifies the DAG by representing all unmeasured variables with the same causal structure (i.e., the same arrows in and out) as a single node [47]. In this case, we included the concept of “familiarity” as that node representing all common determinants of both, place of residence and race. On the other hand, vulnerability (as a context variable measured with the IPVS) in which people are born, would neither determine or be determined by race. But it would be determined by factors such as familiarity and place of residence. The patterns of participation would be a common effect of sex and age, however, even if it was conditioned this would not change the adjustment set suggested for the DAG. A double-sided broken arrow between sex and race was included to represent the hypothesis of an interaction between race and sex in oral health.

Fig 1

Directed acyclic graphs, with the variables and implications tested, for tooth loss according to race/skin color and sex.

The independence implications suggested by the DAG were assessed using statistical tests according to the nature of the dependent variable (e.g., ordinal logistic regression for ordinal outcome; linear regression for continuous variable). The level of significance to reject the testable implications was adjusted according to the Holm-Bonferroni method. The resulting DAG (Fig 1) had seven testable implications of independence, for which 19 tests were performed, considering that some variables were polytomous (adjusted significance level: 0.0026). However, none of the implications were rejected (p>0.10 for all tests), which was interpreted as an indicator of consistency between the DAG and the data [39, 45].

The minimal sufficient adjustment set suggested by the DAG included IPVS and Age. Consequently, adjustment for the unmeasured variables was not necessary. For the variables used, the most functional way to represent their association with the outcome was evaluated. Age was categorized by decile and IPVS remained with the categories of origin. Both variables were consistent and progressively associated with tooth loss (Fig 2), and therefore, included in the model as continuous.

Fig 2

Prevalence of tooth loss by age decile and by IPVS category.

The analyses were conducted using the svy (survey) option of STATA 14.0, which considers the study design weights with complex sampling. In univariable analysis, associations between sociodemographic factors and race were tested using Pearson’s chi-square test, according to sex. The adjusted Prevalence Ratio (PR) and their 95% intervals (95%CI) of tooth loss associations with the variables sex, race, and their interaction were estimated by Poisson regression (robust). Since this is a cross-sectional study, interaction measures were analogous to the Ratio of Relative Risks (RR) and Relative Excess Risk due to Interaction (RERI), which in our study we named Relative Prevalence Ratio (RPR) and Relative Excess Prevalence due to Interaction (REPI) on the multiplicative and additive scale, respectively. These estimates were adjusted by IPVS and Age. The interaction was evaluated on the additive and multiplicative scales, estimating, respectively, the relative excess prevalence due to interaction (REPI), estimated in a similar way to the relative excess risk due to interaction (RERI) calculated in cohort studies, and Relative Prevalence Ratio (RPR, analogous to the ratio of Relative Risks) and their 95%CI [40]. Considering that the event studied occurred after the exposures of interest (race and sex), the prevalence ratio was considered a good approximation of the RR. The results of the interaction analyses were presented according to the recommendations of Knol and VanderWeele [39]. As complementary measures to assess the interaction on the additive scale, we calculated the S-index and the attributable proportion (AP). However, because the denominator of the S-index was negative, of those both, we only reported the AP [48], which corresponds to the ratio between the REPI and the prevalence ratio of the category considered as doubly exposed (black women) [48].

To illustrate the interpretation of the resulting model, the predicted prevalence for each of the four categories defined by race and sex were calculated, taking as a reference the white man and a central tendency value of the adjustment variables.

The distribution of the tooth loss exhibited a significant number of "zeros" (see S1 Fig). Therefore, we decide to focus the primary analysis on the dichotomous outcome. However, to rule out that this decision would bias the conclusions due to loss of information, we also analyzed the outcome as the ordinal variable of tooth loss as aforementioned, to assess the consistency of the trends observed. Consequently, adjusted Odds Ratio were also estimated using multiple ordinal logistic regression models. In this case, four categories of sex and race combinations were created, and the model was also adjusted by age and IPVS. The project was approved by the Research Ethics Committee of the University of Campinas (Opinion no. 3744551/2019 of 04/12/2019; CAAE no. 24860219.4.0000.5404).

Results

After excluding individuals who declared themselves to be indigenous, yellow individuals and another race, the data of 2962 people aged 10 years or more were analyzed. 558 (18.9%) black and 1060 (35.8%) white women took part in the research. Among men, 504 (17.0%) were black and 840 (28.4%), white. There was a trend in the black population for the younger age groups, while for whites this trend occurred for the older age groups. The percentage of women who receive an income greater than 3 MW was more than three times higher for white women than for black women. For schooling, the percentage who studied 9 years or more was 30% higher for white women. Black men also have lower income compared to white ones. The percentage of men living in a place of very low vulnerability was 10 times higher among white men and 3 times higher among black women, when comparing data between categories. (S1 Table).

It was observed that 52% had lost at least one or more teeth, in the total population studied. In the crude analysis, tooth loss of all teeth was more frequent in women, particularly in white women, and less frequent in black men (S1 Fig). After adjustments, the prevalence of tooth loss was 19% higher in black women compared to white men (PR: 1.19; 95%CI: 1.05–1.34). Among the female population, this event was 26% higher in black women compared to white women (PR: 1.26; 95%CI: 1.13–1.42). In the corresponding race category, black women had a 14% higher prevalence of tooth loss (PR: 1.14; 95%CI: 1.02–1.27) compared to black men (Table 1).

Table 1

Adjusted Prevalence Ratios (PRs) for tooth loss comparing sex and race groups (ISACamp 2014/15).

	Tooth Loss				PRs (95% CI) for race within strata of sex
	White Individuals		Black Individuals
	Cases/ non-cases (Prevalence; IC95%)	PR (95%CI)	Cases/ non-cases (Prevalence; IC95%)	PR (95%CI)
Men	439/400	1	206/297	1.04 (0.88–1.23)	1.04 (0.88–1.23)
	(52.3%; 48.9%-55.8%)		(41%; 36.6%-45.4%)	p = 0.61	p = 0.61
Women	597/462	0.94 (0.84–1.06)	297/261	1.19 (1.05–1.34)	1.26 (1.13–1.42)
	(56.4%; 53.3%-59.4%)	p = (0.32)	(53.2%; 49%-57.5%)	p = 0.006	p < 0.001
PRs (95% CI) for sex within strata of race		0.94 (0.84–1.06)		1.14 (1.02–1.27)
		p = 0.32		p = 0.021

REPI (95%CI) = 0.20 (0.04–0.37) p = 0.02.

Relative Prevalence Ratio (RPR) (95%CI) = 1.21 (1.01–1.44); p = 0.04.

PRs are adjusted by IPVS and age, and weighted according to complex sample design.

Thus, race and sex presented an interaction both on the additive and on the multiplicative scale (Table 1). In black women, the prevalence of tooth loss was 20% higher than expected (95%CI: 0.04–0.37; p = 0.02) on the additive scale; and 21% higher than expected (RPR = 1.21; 95%CI: 1.01–1.44; p = 0.04) on the multiplicative scale. According to the AP, 17.17% (95%CI: 2.5% - 31.86%, p = 0.02) of the prevalence in black women is due to the interaction. Fig 3 shows the estimated values for each of the categories defined by race and sex, taking as mean parameters of IPVS equal to 2 and age of 37 years. With these parameters, the tooth loss calculated according to the adjusted PR for black women was 45.1% (95%CI: 40.2%– 50.0%), which was higher than the values expected on the additive, 37.4% (95%CI: 29.8%– 44.7%) and multiplicative scales, 37.3% (95%CI: 29.7%– 45.0%).

Fig 3

Percentage of tooth loss estimated according to the prevalence ratio versus expected on the multiplicative and additive scales, according to the categories of race and sex adjusted by IPVS, ISACamp 2014/15.

Using ordinal logistic regression comparing tooth loss across the races and sex groups, adjusted for age and IPVS, it was observed that black women have twice as much odds of passing to a category of higher tooth loss (OR = 2.10: 95%CI = 1.52–2.91: p<0.001), compared to white men (Table 2). In this model, no other category was significantly different from the group of white men.

Table 2

Odds ratio of tooth loss for the oral health outcome between sexes and race in the population with the IPVS and age conditions, ISACamp 2014/15.

Characteristic related to Oral Health (Tooth Loss)	ORa,b (95%CI)	P-value
Black Men	1.15(0.78–1.70)	0.49
White Women	0.92(0.69–1.22)	0.56
Black Women	2.10(1.52–2.91)	<0.001

White Men was the reference category.

Ordinal Logistic Regression adjusted by IPVS and Age.

Discussion

We have shown an interaction between sex and race on the risk for tooth loss on both the additive and the multiplicative scales. This result pointed to a higher prevalence of tooth loss among black women. The analysis of the interaction between race and sex in tooth loss was investigated for the first time on a based-population approach. In this study, information on toothlessness were collected by self-reports; however, self-reported tooth counting has proven to be a validated and reliable indicator of tooth loss [49, 50].

Some studies have found greater tooth loss in women [51-53]. According to a national epidemiological survey [54], between 15 and 19 years old, tooth loss is 20% in females and 13% in males. In adulthood, between 35 and 44 years old, estimates point to 21% of tooth loss in women and 18% in men. In the elderly, the percentage of edentulism rises to 56% in women, against 49% in men. Other authors have also found that the oral health condition in elderly women is worse compared to that of men [31].

Although there is a consensus on the disadvantage of women in oral health, information indicates that the female population seeks dental care more than men [11, 55], as it also happens for other types of health care [11]. This fact may be due to the greater self-perception of health that women have compared to men, in addition to the trend towards greater health care [4, 55]. However, there is evidence that the increased demand for dental treatment does not necessarily reflect better oral health conditions in women [52].

The findings of this study corroborate the results of a study carried out in Brazil, which evaluated the relationship between black individuals and tooth loss [56]. Although the mentioned study refers to the elderly population, the authors found that race is a limiting factor in access to oral health services, with black people being twice as likely to having never gone to the dentist compared to white people. In another Brazilian study, Souza et al. [57] found that, even after adjustments for income and living conditions, self-assessment of oral health, of the care received in dental consultations as bad, and consultation with the aim of extraction are more prevalent oral health issues in black people compared to white. The authors discuss issues of racial segregation beyond socioeconomic status.

Some studies [13, 58] lead to the discussion that being black individuals is a factor that interferes in the decision-making process to recommend tooth extraction. In addition, blacks suffered more posterior teeth extraction when compared to other ethnic groups [24]. It is possible that the dentist’s decision has a pro-white bias to recommend a more complex treatment, more conservative of the tooth, while for black patients in the same condition, extraction is recommended [13, 59]. These tooth extraction practices may also be due to the lower socioeconomic level among the black population, which makes treatments to preserve teeth less accessible. Another explanation for radical decisions would be the issues of prejudice and discrimination, as has been discussed in other studies [13, 57, 58].

A possible explanation for the interaction between race and sex in tooth loss, with a disadvantage for black women, would be the health service conditions with a greater tendency to tooth extraction in black people [13, 58]. If the search and frequency of oral health services are higher among the female population [55], the burden of mutilating practices comes on the black woman.

Although progress has been made to reduce socioeconomic inequalities in the provision of public health services, it is worth noting that in oral health these disparities still tend to persist [23]. Currently, Brazil has some affirmative action policies, such as the quota system, in universities and in public tendering, to repair injustices against black individuals, indigenous, and the poor people. In addition, the National Policy for the Comprehensive Health of the Black Population was created to fight institutional racism and ethnic-racial prejudice in health institutions, instructing social movements, professionals, and health managers to provide better quality health care for the black population [60]. However, what can be verified is that these isolated policies are not able to avoid culturally rooted racism, especially when it comes to oral health services, which are typically for the elite.

Before these results, it is worth noting the need for public policies with affirmative actions and educational measures that can benefit black women to be maintained and rethought, to ensure universal and equal access, as advocated by the Single Health System (In Brazil, called SUS). In addition, this study also draws attention to the need for further studies on racial and sex disparities in oral health and dental issues occupying more space in government policies, to minimize the losses caused by difficulties and inequities.

The strengths of the study come from the novelty in evaluating the interaction between race and sex in tooth loss. Moreover it is a population-based study obtaining a probabilistic sample designed to represent a municipality with around one million inhabitants, with a questionnaire that allows the measurement of relevant variables for the control by potential confounding phenomena. In addition, a careful process of variable selections was made to control confounding. The DAG represented the structures of causal phenomena, which helped to choose the set of minimum variables for valid adjustment. Thus, it avoided unnecessary adjustments overfitting and collision biases that would affect the validity of the estimates. Thus, the association measures found, despite the design limitation, can be considered robust to describe the relationship between sociodemographic factors and to estimate the interaction between race and sex on the outcome of tooth loss.

The limitation of the study are that tooth loss was coded according to the question of the survey, and it was not possible to use the international classification of functional dentition. However, regarding the categories analyzed, as described in S1 Fig, the outcome had an important proportion of people without tooth loss suggesting that a valid and informative approach is to focus on differentiating loss and no loss. Moreover, we also considered the outcome as an ordinal variable resulting in consistent associations. It is also worth considering the survival bias, since we are dealing with tooth loss; adult and elderly individuals are more exposed to this condition. Therefore, this bias could affect the similarity between the prevalence ratio and the relative risk. However, the prevalence measure obtained portrays the current disease burden and may indicate inequities in health between the groups analyzed.

Conclusion

This study found a significant interaction between race and sex in tooth loss, with a disadvantage for black women. The female population who that self declared black was twice odds of having a worse dental condition than white men. The results could indicate a trend in oral health care based on mutilating decision-making in black individuals, affecting, especially, black women. Our findings suggest attention to the distribution pattern of tooth loss by race and sex and more studies can confirm the consistency between the interactions that indicate greater vulnerability in black women. Thus, after the evidence found in this work, it is recommended reinforce the surveillance to recognize and eliminate inequities in oral health care.

Sociodemographic characteristics according to sex and race, in the population aged 10 or over, ISACamp 2014–2015.

(DOCX)

Click here for additional data file.

Percentage of frequency of tooth loss, according to race, by sex, ISACamp 2014/15.

(DOCX)

Click here for additional data file.

13 Jan 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Mariana Minatel Braga

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Revision PONE-D-21-32428

Race (black-white) and sex inequalities in tooth loss: a population-based study

1. The authors do not explain the concept of intersectionality in the introduction and do not justify the use of race in biomedical publications. Authors are based on too old references on the topic.

2. The RERI analysis is sound. However, without a proper justification for the outcome ( lost of 1 tooth or no – What is the usefulness? ) and a miscellaneous in ordinal category ( No loss, loss of 1 tooth, loss of more than one tooth, or loss of all teeth) in complicated to readers understand.

3. The DAD used by the authors is difficult to understand. Authors do not put "Place of residence" or "vulnerability index" linked to race. This is a mistake due to the structural factors related to ethnic inequalities. Recent evidence explaining racial inequities put it on evidence (neighbourhood conditions) in oral health (https://pubmed.ncbi.nlm.nih.gov/29634429/). Moreover, the authors did not analyze health behaviours, important mediators explaining tooth loss.

4. The outcome of tooth loss is very strange for me. For population comparisons, the authors do not use international classifications like functional dentition (more than 20 permanent teeth) and severe tooth loss ( less than ten remaining teeth).

5. I do not understand why authors opted to maintain ten years (children with mixed dentition) or more in the sample. Joining children, adolescents, adults and older adults in the same analysis. Difficult to interpret

6. Descriptive analysis need the n and CI 95% ( supplemental file 1)

7. How was tooth loss collected in the sample? By the number of teeth? Self-report one or more than one? Need more explanation…

8. Need more information on how authors collected data. How many examiners? Calibration? Provide detailed information.

9. Did the author proceed to sensitivity analysis? "It was verified that the categories of black and brown presented a similar association for tooth loss, in each sex" …and age between browns and blacks? Were they similar or different from Whites?

10. Why did the RERI analysis not report the attributable proportion [AP] and synergy index [S]?

11. Tables need to show the n of each category and prevalence and 95% CI, as recommended by Knol & VanderWeele (2012). This will add additional transparency for readers

12. Analyzing absolute prevalence in tooth loss, the percentages (Table 1 ) for white men is 52% ( authors did not put 95% CI), for white women 59% ( why prevalence ratio (PR)under 1?), black man 41% ( lower than white man? ) and black women 53% ( lower than white woman?). ….with slight absolute differences. Problems with the outcome selection as commented in comments #2 and #5 joining children with older adults…

Reviewer #2: The topic studied is relevant and well-justified in the Introduction section.

Methods are well-reported and properly designed to answer the research question. A few typos should be revised (e.g., Ln. 92 – the classification “very low” is presented twice).

Data were analyzed adequately, and the results are clearly presented and discussed.

Ln. 178 – Supplementary Figure 1 shows the opposite behavior.

Reviewer #3: I would like to congratulate the authors for their work and willingness to investigate differences between racial minorities and their privileged peers in an outcome that is of great importance in the field of public health. I'm sure the authors will appreciate my comments aimed at providing an outside view on how this manuscript can improve.

Introduction

- I would like to ask authors not to use "Blacks". Please use these words only as adjectives, using them as a noun is dehumanizing. Please write "Black individuals", for example.

- In the introduction, the authors present that the main health inequities related to race are: a) due to lesser access to health services; and b) due to socioeconomic issues. However, these are just two of the many factors where race is a cause of health inequities. It is not correct to reduce or omit other factors such as the slavery past of Brazilian society and Latin America. Thus, considering that the authors aim to address the influence of race on tooth loss, it is necessary that the authors deepen the discussion on this topic in the introduction.

- Please provide a very specific reason for what race means and why it is important to study it in the breakdown of oral health outcomes.

- I suggest that the authors review the objectives of the study, as in the format it is written, the authors state that there is an association between race and tooth loss. I suggest splitting it into two objectives: a) to investigate a possible association between race and tooth loss and b) assess possible interactions between race and sex in association with tooth loss.

There is also a need for greater attention to standardize the words used throughout the text, such as “race”, “race/skin color”, and “self-reported skin color” to refer to different racial groups in Brazil. In addition to needing to clearly answer, right in the introduction, why to study racial and gender inequities in face of tooth loss, the authors need to review the use of “skin color” throughout the text. If the authors are referring to the Brazilian classification system, it is important to make this clear, and to recognize first that the official wording has been "color/race" since 1991, and probably the same one used in the study. It should be clear that the complexity of the classification is beyond skin color and much more related to how society treats racial minorities, given racism and racial discrimination. I suggest that authors choose only "race" to facilitate understanding by international readers, making this reduction clear from the official classification. Please provide a very specific reason for what race means and why it is important to study it in the breakdown of oral health outcomes.

- The authors were not able to clearly show in the introduction how the current manuscript has originality compared to other articles already published.

- In addition, after reading the introduction, I would like to suggest some references for the authors to read before proceeding with the reformulation of the manuscript:

- Phelan JC, Link BG. Is Racism a Fundamental Cause of Inequalities in Health? Annu Rev Sociol. 2015;41(1):311-330.

- Bastos JL, Celeste RK, Paradies YC. Racial Inequalities in Oral Health. J Dent Res. 2018:22034518768536.

- Costa F,et al., Racial and regional inequalities of dental pain in adolescents: Brazilian National Survey of School Health (PeNSE), 2009 to 2015. 2021 Cad Saude Publica. 2021 Jun 25;37(6):e00108620.

- Williams DR, Priest N, Anderson NB. Understanding associations among race, socioeconomic status, and health: Patterns and prospects. Health Psychol. 2016;35(4):407-411.

Methods:

- In the section of study variables, please, provide all cathegorization of all variables. For example, In which category were yellow and indigenous individuals included?

- I recommend changing the label from the "black" to "black/brown individuals" since brown individuals were also included in this category.

- Please also present a paragraph supporting the assumptions presented in the DAG.

- how was tooth loss originally collected by DMF-T?

Discussion:

- Line 265, page 10. “SUS” needs to be written out in full.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: Yes: Luiz Alexandre Chisini

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

12 Mar 2022

PONE-D-21-32428

Race (black-white) and sex inequalities in tooth loss: a population-based study

PLOS ONE

Dear editors

We are submitting the new version of our manuscript. We would like to thank the reviewers for the suggestions, which have certainly enriched our paper. All comments and suggestions were fully taken into account in the preparation of this new version. Our point-by-point responses to the issues raise are listed below:

Review Comments to the Author

Reviewer #1:

1. The authors do not explain the concept of intersectionality in the introduction and do not justify the use of race in biomedical publications. Authors are based on too old references on the topic.

Response: We modified the text and added the concept of intersectionality in the introduction: “In this way, it is worth considering the analytical strategy of intersectionality, developed by Crenshaw in 1989 to study the black feminism. Intersectionality, drawing on a set of social issues and perspectives, explain an experience of multiple subordination that cannot be reduced to the simple sum of underprivileged groups [34]”. (Lines 59-63).

We also justify the use of race in biomedical publications: “Racial minorities, in this case black people, may biologically embody the effects of racism [7,12], with everyday (or even less common) discriminatory exposures [12]. Adversities throughout life, such as poverty, psychosocial stresses, stereotypes, and the context of housing, can affect the physical and mental health, altering cardiocirculatory, metabolic and immunological functions [8]. In oral health is possible that inequities are due to poverty [10], levels of education [4] or discrimination in health care [13]. In Brazil, skin color has been studied as proxy of race and this indicator is commonly referred as color/race [14,15,16]. In health, in contrast only to the biologic approach, race may be considered as a concept socially constructed by historic dynamics and power relations [8,12].” (Lines 27-34)

2. The RERI analysis is sound. However, without a proper justification for the outcome (lost of 1 tooth or no – What is the usefulness? ) and a miscellaneous in ordinal category (No loss, loss of 1 tooth, loss of more than one tooth, or loss of all teeth) in complicated to readers understand.

Response: Thank you for the comment and I take the opportunity to clarify some points. We added this justification on the text: “Tooth loss is an important marker of oral health, since each tooth is a dental organ and its subtraction has consequences on occlusal adjustment, temporomandibular joint disorder, reduction of the masticatory board and consequently overload for the digestive system [37]. Besides, the loss of at least one tooth since it represents neglect in the dental field, resulting in an increase in the level of severity of oral diseases, in addition to reflecting the model of oral health care adopted and the way individuals understand the condition [38].” (Line 97-103)

Moreover, in the last paragraph of methodology, we complemented: “… Therefore, we decided to focus the primary analysis on the dichotomous outcome, which also represents the most practical way to assess and present interactions [39, 40]. However, on the other hand, the analysis of the same variable as ordinal can show us a gradient of tooth loss, varying its degrees of loss [41]. This analysis was used only to confirm trends in tooth loss in the black population. (Lines 103-106)

3. The DAD used by the authors is difficult to understand. Authors do not put "Place of residence" or "vulnerability index" linked to race. This is a mistake due to the structural factors related to ethnic inequalities. Recent evidence explaining racial inequities put it on evidence (neighbourhood conditions) in oral health (https://pubmed.ncbi.nlm.nih.gov/29634429/). Moreover, the authors did not analyze health behaviours, important mediators explaining tooth loss.

Response: We appreciate the comments and the reference indicated and we reviewed the DAG. We understand that neighborhood conditions are important aspects in the health and specifically in oral health. But our thinking about the DAG is the issues of determination. We understand that race is closely linked to places of residence, but these variables are not directly related in a causal diagram. Explicitly, a race does not cause a place of residence, nor a place of residence cause a race. These variables can be associated, as the reviewer correctly indicated, but the only explanation it would be sharing a common cause.

Because can be a long list of potential unmeasured variables, fitting this role of being a common cause of place and race, we follow a common convention that visually simplifies the DAG by representing all unmeasured variables with the same causal structure (i.e., the same arrows in and out) as a single node (Digitale JC, Martin JN, Glymour MM. Tutorial on directed acyclic graphs. J Clin Epidemiol. 2021 Aug 8:S0895-4356(21)00240-7. doi: 10.1016/j.jclinepi.2021.08.001). In this case, we included the concept of “familiarity” as that node representing all common determinants of both, place of residence and race.

In the same sense, vulnerability (as a context variable measured with the IPVS) in which people are born, would neither determine nor be determined by race. But it would be determined by factors such as familiarity and place of residence.

On the other hand, we agree with reviewer regarding the health behaviours as important mediators of demographic determinants. Therefore, we included a variable on the DAG (health behavior) determined by sex, race and age, which would be causally connected with the outcome (Please see new DAG with this inclusion). However, such our objective was to estimate the total effects (not to analyse mediation) of race and sex, specifying their interaction, no additional adjustment was necessary.

During revision of DAG, we perceived that pattern of participation would be better described as a common effect of sex and age than a common cause. So, as it can be conditionate by decision of participate of population, it can imply a collider bias, opening a path. However, this bias it would be also corrected by adjusting by age. Therefore, again, the set of variables indicated to adjusted remain the same. Inserted in the text lines 149-161.

4. The outcome of tooth loss is very strange for me. For population comparisons, the authors do not use international classifications like functional dentition (more than 20 permanent teeth) and severe tooth loss (less than ten remaining teeth).

Response: The question used to collect the tooth loss data used as answered in question #7 and added to the text used the categories (1) No, (2) yes, only one tooth, (3) yes, more than one tooth, (4) yes, all teeth, being used in this way in other studies by Gomes Filho and Verçosa. We insert an information on the studies limitations: “Also, tooth loss was coded according to question of the survey, and it was not possible to use the international classification of functional dentition.” (Lines 323-324).

5. I do not understand why authors opted to maintain ten years (children with mixed dentition) or more in the sample. Joining children, adolescents, adults and older adults in the same analysis. Difficult to interpret

Response: We considered important to analyze a wide age range, since we believe that inequality problems affect the population from early stages of life. In this sense, we observed that the associations were consistent and independent of age in the multiple model. We understand the reviewer's concern, however, as we described in the methods, our interviewers were trained and instructed to ask the question about tooth loss excluding teeth extracted for orthodontic reasons, disregarding teeth extracted for orthodontic, wisdom and baby tooth reasons (this information has been added to the text). This can minimize the risk of a classification bias that the age group would have in relation to studies of tooth loss and social conditions. (Lines 96-97).

6. Descriptive analysis need the n and CI 95% (supplemental file 1)

Response: Thanks for the suggestion, the data was entered into the table. (Page 17)

7. How was tooth loss collected in the sample? By the number of teeth? Self-report one or more than one? Need more explanation…

Response: We added a information in the text: “The outcome variable was self-referred tooth loss, extracted from the question: “Have you ever lost a tooth (upper or lower)? If so, did he lose one, more than one or all of his teeth? (disregard extracted teeth to place braces, wisdom and milk teeth)”, with the options to response: (1) No, (2) yes, only one tooth, (3) yes, more than one tooth, (4) yes, all teeth.” (Lines 90-93)

We also entered the information that the tooth loss was self-reported: “The analysis of the interaction between race and sex in tooth loss was investigated for the first time on a population basis. In this study, information on toothlessness were collected by self-reports…” (Lines 261-262)

8. Need more information on how authors collected data. How many examiners? Calibration? Provide detailed information.

Response: We added the following information: “Data were collected from 3021 people aged 10 or over [36], using a pre-coded questionnaire, applied through tablets, by interviewers trained." (Lines 85-87). “When asking about tooth loss, the interviewers informed the respondents to exclude teeth extracted for orthodontic reasons, disregarding teeth extracted for braces, wisdom and baby teeth.” (Lines 96-97)

9. Did the author proceed to sensitivity analysis? "It was verified that the categories of black and brown presented a similar association for tooth loss, in each sex" …and age between browns and blacks? Were they similar or different from Whites?

Response: We inform that it was found that the age categories for blacks and browns showed a similar association for tooth loss. This information was inserted in the text: “ It was verified that the categories of black and brown presented a similar association for tooth loss, in each sex and in each age, so these two categories were joined to compose the group of black individuals (Also used by IBGE-Brazilian Institute of Geography and Statistics) [42]. ” (Lines 110-112).

10. Why did the RERI analysis not report the attributable proportion [AP] and synergy index [S]?

Response: We appreciate the comment and added the suggested analyzes. In methodology, we included: “As complementary measures to assess the interaction on the additive scale, we calculated the S-index and the attributable proportion (AP). However, because the denominator of the S-index was negative, of those both, we only reported the AP [48], which corresponds to the ratio between the REPI and the prevalence ratio of the category considered as doubly exposed (black women) [48]”. (Lines 188-192). Moreover, in result section, we reported that “According to the AP, 17.17% (95%CI: 2.5% - 31.86%, p=0.02) of the prevalence in black women is due to the interaction”. (Lines 233-234).

11. Tables need to show the n of each category and prevalence and 95% CI, as recommended by Knol & VanderWeele (2012). This will add additional transparency for readers.

Response: We appreciate the recommendation. We reviewed the manuscript and verified that the table already followed exactly the orientations of the reference. That is, the number of cases and non-cases in each category, the association measure and its confidence interval. However, we agree that it would be good to report unadjusted prevalence. Therefore, we added the prevalence with its confidence interval in each category. (Page 9)

12. Analyzing absolute prevalence in tooth loss, the percentages (Table 1 ) for white men is 52% ( authors did not put 95% CI), for white women 59% ( why prevalence ratio (PR)under 1?), black man 41% ( lower than white man? ) and black women 53% ( lower than white woman?). ….with slight absolute differences. Problems with the outcome selection as commented in comments #2 and #5 joining children with older adults…

Response: Different than unadjusted measures, the estimates of prevalence ratios in this manuscript considered both the sample weights of the complex design of the study and the adjustment for age and IPVS, as informed in the text: “The analyses were conducted using the svy (survey) option of STATA 14.0, which considers the study design weights with complex sampling”. (Lines 177-178). We appreciate the comment, and we agree. So, we added the explanation on the footnote of the table: “PRs are adjusted by IPVS and age, and weighted according to complex sample design”

We reviewed the calculations and confirmed the values. (Page 9).

Reviewer #2:

The topic studied is relevant and well-justified in the Introduction section.

Methods are well-reported and properly designed to answer the research question. A few typos should be revised (e.g., Ln. 92 – the classification “very low” is presented twice).

Response: We appreciate the recognition and inform you that the suggestions have been modified in the text. (Line 129).

Data were analyzed adequately, and the results are clearly presented and discussed.

Ln. 178 – Supplementary Figure 1 shows the opposite behavior.

Response: In the crude analysis, black men have the lowest tooth loss and among women, black women have less tooth loss compared to white women. However, when we do the adjusted analysis, we find the opposite.

As explained in the text: “In the crude analysis, tooth loss of all teeth was more frequent in women, particularly in white women, and less frequent in black men (Supplementary Figure 1). After adjustments, the prevalence of tooth loss was 19% higher in black women compared to white men (PR: 1.19; 95%CI: 1.05-1.34).” (Lines 217-220)

Reviewer #3:

I would like to congratulate the authors for their work and willingness to investigate differences between racial minorities and their privileged peers in an outcome that is of great importance in the field of public health. I'm sure the authors will appreciate my comments aimed at providing an outside view on how this manuscript can improve.

Response: We appreciate the recognition.

Introduction

1. I would like to ask authors not to use "Blacks". Please use these words only as adjectives, using them as a noun is dehumanizing. Please write "Black individuals", for example.

Response: We changed for “black individuals”.

2. In the introduction, the authors present that the main health inequities related to race are: a) due to lesser access to health services; and b) due to socioeconomic issues. However, these are just two of the many factors where race is a cause of health inequities. It is not correct to reduce or omit other factors such as the slavery past of Brazilian society and Latin America. Thus, considering that the authors aim to address the influence of race on tooth loss, it is necessary that the authors deepen the discussion on this topic in the introduction.

Response: We appreciate the suggestion and inform you that we have modified the text as suggested. (Lines 25-34).

3. Please provide a very specific reason for what race means and why it is important to study it in the breakdown of oral health outcomes.

Response: Thank you. We have included this information in this paragraph of the text: “Racial minorities, in this case black people, may biologically embody the effects of racism [7,12], with everyday (or even less common) discriminatory exposures [12]. Adversities throughout life, such as poverty, psychosocial stresses, stereotypes, and the context of housing, can affect the physical and mental health, altering cardiocirculatory, metabolic and immunological functions [8]. In oral health is possible that inequities are due to poverty [10], levels of education [4] or discrimination in health care [13]. In Brazil, skin color has been studied as proxy of race and this indicator is commonly referred as color/race [14,15,16]. In health, in contrast only to the biologic approach, race may be considered as a concept socially constructed by historic dynamics and power relations [8,12].”

4. I suggest that the authors review the objectives of the study, as in the format it is written, the authors state that there is an association between race and tooth loss. I suggest splitting it into two objectives: a) to investigate a possible association between race and tooth loss and b) assess possible interactions between race and sex in association with tooth loss.

Response: We accept the suggestion, and the text was changed lines 64-65.

5.There is also a need for greater attention to standardize the words used throughout the text, such as “race”, “race/skin color”, and “self-reported skin color” to refer to different racial groups in Brazil. In addition to needing to clearly answer, right in the introduction, why to study racial and gender inequities in face of tooth loss, the authors need to review the use of “skin color” throughout the text. If the authors are referring to the Brazilian classification system, it is important to make this clear, and to recognize first that the official wording has been "color/race" since 1991, and probably the same one used in the study. It should be clear that the complexity of the classification is beyond skin color and much more related to how society treats racial minorities, given racism and racial discrimination. I suggest that authors choose only "race" to facilitate understanding by international readers, making this reduction clear from the official classification. Please provide a very specific reason for what race means and why it is important to study it in the breakdown of oral health outcomes.

Response: Thank you. We used “race” to define “color/race” and we clarify in the methods section (lines 107-110): “The race variable studied was self-reported. In Brazil this is the official method of racial classification since 1991, based on the individuals’ own perception of the color of their skin. In the present study, race/skin color variable was denominated “race” and categorized as black, brown, and white individuals.”.

In the introduction, we also added an information of use of race as a social construct: “In health, in contrast only to the biologic approach, race may be considered as a concept socially constructed by historic dynamics and power relations [8,12].” (Lines 33-34)

After a paragraph explaining about gender inequalities in tooth loss (lines 59-63)., we include the need and possibility of studying race and gender in intersectional aspects: “The evidence of racial and gender inequalities in tooth loss could bring the subgroup of black women underprivileged in relation to this theme. In this way, it is worth considering the analytical strategy of intersectionality, developed by Crenshaw in 1989 to study the black feminism. Intersectionality, drawing on a set of social issues and perspectives, explain an experience of multiple subordination that cannot be reduced to the simple sum of underprivileged groups [34].”

6. The authors were not able to clearly show in the introduction how the current manuscript has originality compared to other articles already published.

Response: We thank you for the note made, to make the originality of the study clear, we added: “To our knowledge, there are no studies looking at the relationship between race and sex in relation to tooth loss and this study contributes to filling this gap.” (Lines 66-67).

7. In addition, after reading the introduction, I would like to suggest some references for the authors to read before proceeding with the reformulation of the manuscript:

-rdszz Phelan JC, Link BG. Is Racism a Fundamental Cause of Inequalities in Health? Annu Rev Sociol. 2015;41(1):311-330.

- Bastos JL, Celeste RK, Paradies YC. Racial Inequalities in Oral Health. J Dent Res. 2018:22034518768536.

- Costa F,et al., Racial and regional inequalities of dental pain in adolescents: Brazilian National Survey of School Health (PeNSE), 2009 to 2015. 2021 Cad Saude Publica. 2021 Jun 25;37(6):e00108620.

- Williams DR, Priest N, Anderson NB. Understanding associations among race, socioeconomic status, and health: Patterns and prospects. Health Psychol. 2016;35(4):407-411.

Response: We are very grateful for the suggestions and inform you that they were fundamental in making our introduction more grounded.

Methods:

8. In the section of study variables, please, provide all cathegorization of all variables. For example, In which category were yellow and indigenous individuals included?

Response: We added an information about yellow and indigenous individuals: “Individuals who declared themselves as yellow, indigenous, or other races were not studied, as they represented less than 2% of the population.” (Lines 113-114).

9. I recommend changing the label from the "black" to "black/brown individuals" since brown individuals were also included in this category.

Response: We made the change in the text. (Lines 110-111)

10. Please also present a paragraph supporting the assumptions presented in the DAG.

Response: We made the change in the text line 146. “In the DAG, we represented the effects of race and sex on oral health, which can be mediated by income [3,4], education [4] and, more proximally to the outcome, by health behaviors and the dental care [6]. Other variables, such as age, familiarity, patterns of participation in the research, place of residence and IPVS [7], were also considered in the causal diagram (Figure 1). Because it can be a list of unmeasured common causes of both place and race, we follow a convention that visually simplifies the DAG by representing all unmeasured variables with the same causal structure (i.e., the same arrows in and out) as a single node (47). In this case, we included the concept of “familiarity” as that node representing all common determinants of both, place of residence and race. On the other hand, vulnerability (as a context variable measured with the IPVS) in which people are born, would neither determine or be determined by race. But it would be determined by factors such as familiarity and place of residence. The patterns of participation would be a common effect of sex and age, however, even if it was conditioned this would not chance the adjustment set suggested for the DAG. A double-sided broken arrow between sex and race was included to represent the hypothesis of an interaction between race and sex in oral health.” (Lines 149-161)

- how was tooth loss originally collected by DMF-T?

Response: We added an information about the original question about tooth loss: “The outcome variable was self-referred tooth loss, extracted from the question: “Have you ever lost a tooth (upper or lower)? If so, did he lose one, more than one or all of his teeth? (disregard extracted teeth to place braces, wisdom and milk teeth)”, with the options to response: (1) No, (2) yes, only one tooth, (3) yes, more than one tooth, (4) yes, all teeth.” (Lines 90-93)

We also entered the information that the tooth loss was self-reported: “The analysis of the interaction between race and sex in tooth loss was investigated for the first time on a population basis. In this study, information on toothlessness were collected by self-reports; ..” (Lines 261-262)

Discussion:

11. Line 265, page 10. “SUS” needs to be written out in full.

Response: We appreciate your suggestion and inform you that it has been added to the text line 306.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

8 Aug 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Mariana Minatel Braga

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: No

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: All suggestions done by this reviewer in the first version were adequately addressed. Moreover, accepting the recommendations done by the other reviewer substantially improved the revised version of the manuscript.

Reviewer #3: The present study has improved considerably in this version. I still have a few points to point out.

Skin color self-report should not be a limitation of the study, as this is the best strategy to investigate race in Brazil, since race and skin color can be interpreted as synonyms in the Brazilian context.

In addition, rethinking from the statements of reviewer #1, I believe that the inclusion of individuals with mixed dentition is really difficult to justify and can imply large biases as well as the way of categorizing tooth loss. The authors mention that they used the question of the study by Gomes Filho and Verçosa, however, they collected the answers differently. This is a critical point of the study.

The authos relate “After adjustment for covariates suggested by a directed acyclic graph”; however, is not appropriate describe that DAG “suggest” adjustments. Report only that DAG was used to select covariates.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: No

Reviewer #3: Yes: Luiz Alexandre Chisini

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

16 Aug 2022

Dear editors

We are submitting the new version of our manuscript. We would like to thank the reviewers for the suggestions, which have certainly enriched our paper. All comments and suggestions were fully taken into account in the preparation of this new version. Our point-by-point responses to the issues raise are listed below:

Review Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

Reviewer #3: All comments have been addressed

Response: We appreciate the comments.

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

Reviewer #3: Yes

Response: We appreciate the comments.

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

Reviewer #3: Yes

Response: We appreciate the comments.

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: No

Reviewer #3: Yes

Response: The raw data and a codebook were deposited in the public repository as requested. Access address: https://doi.org/10.25824/redu/J9YOTW .

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

Reviewer #3: Yes

Response: We appreciate the comments.

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: All suggestions done by this reviewer in the first version were adequately addressed. Moreover, accepting the recommendations done by the other reviewer substantially improved the revised version of the manuscript.

Response: We appreciate the comments.

Reviewer #3: The present study has improved considerably in this version. I still have a few points to point out.

Skin color self-report should not be a limitation of the study, as this is the best strategy to investigate race in Brazil, since race and skin color can be interpreted as synonyms in the Brazilian context.

Response: We agree with the reviewer. We removed the self-declaration race/ skin color from the limitation.

In addition, rethinking from the statements of reviewer #1, I believe that the inclusion of individuals with mixed dentition is really difficult to justify and can imply large biases as well as the way of categorizing tooth loss. The authors mention that they used the question of the study by Gomes Filho and Verçosa, however, they collected the answers differently. This is a critical point of the study.

Response: The interviewers did not consider tooth loss those extracted for orthodontic reasons, disregarding teeth extracted for braces, wisdom, and milk (baby) teeth (lines 105-106). Further, regarding the categories analyzed, as described in figure S1, the outcome had an important proportion of people without tooth loss suggesting that a valid and informative approach is to focus on differentiating loss and no loss. Moreover, we also considered the outcome as an ordinal variable resulting in consistent associations (as commented in the discussion, lines 330-336). Therefore, no additional modifications were made regarding this point.

The authos relate “After adjustment for covariates suggested by a directed acyclic graph”; however, is not appropriate describe that DAG “suggest” adjustments. Report only that DAG was used to select covariates.

Response: We appreciate the suggestion has been modified as recommended. Line 18.

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: No

Reviewer #3: Yes: Luiz Alexandre Chisini

Response: We appreciate the comments.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

29 Sep 2022

Race (black-white) and sex inequalities in tooth loss: a population-based study

PONE-D-21-32428R2

Dear Dr. TERRA e SOUZA,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Mariana Minatel Braga

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Considering the improvements in the revised version and/or adequate answers to reviewers' queries, this manuscript could be, at the present format, acceptable for publication. We appreciate authors' efforts in improving this final version of the manuscript.

5 Oct 2022

PONE-D-21-32428R2

Race (black-white) and sex inequalities in tooth loss: a population-based study

Dear Dr. Terra e Souza:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Mariana Minatel Braga

Academic Editor

PLOS ONE